When I started out in hockey blogging with my co-blogger, Carolyn, one of us understood stats very well, and one of us did not. Carolyn dragged me kicking and screaming (and sometimes crying because, I’ll be honest, trying to understand these made me feel dumb, and I hate feeling dumb) into a basic understanding of what is actually up with analytics.
They say one of the best ways to learn is to teach, so I figured, if I can understand these terms well enough to explain them to someone else, I’ve probably got it down, right? We also figured that if you’re already not mathematically minded (which I’m totally not), it might be easier to have someone ELSE of the liberal arts persuasion try explaining things.
For your basic definitions:
Shot aka Shot on Goal: a shot that would be a goal without the presence of a goaltender
Missed shot: a shot that would not have resulted in a goal regardless of the presence of a goaltender
Blocked shot: a shot blocked by a player other than the goaltender
Slightly more advanced:
Corsi: Corsi is just shots on goal + missed shots + blocked shots. Corsi is used as a measure of possession, as you have to have the puck in your team’s possession to have shots at goal. Corsi can be expressed in several ways. The first is in whole numbers, or Corsi Events.
Say the Stars were playing the Maple Leafs, the Stars had 16 Corsi Events in one period and the Maple Leafs had ten. That means the Stars had 16 shots/missed shots/blocked shots to the Leafs’ ten. In percentages, the Stars had 62% of the Corsi Events, or 62% Corsi For (CF).
(If you don’t remember how to get percentages, you take the Corsi events for the Stars number and divide it by the total number of Corsi events. 16 ÷ 26 = .615, or 62%.)
You can also see Corsi talked about as a Corsi Differential. In the above example, the Stars would be +6 in Corsi differential, since they had 16 events to the Leafs’ ten.
Corsi is measured by counting whoever is on the ice for a shot attempt, regardless of who takes the shot. If Jordie Benn is said to have 62% CF in a game, it means that while he was on the ice, the Stars had 62% of the total Corsi events. Over a season or a series of games, it can be good for measuring how useful a player is to team possession.
Fenwick: Basically, Corsi minus blocked shots. Corsi came first, and some brave soul named Fenwick asked why blocked shots were included in the count. Thus, Fenwick %. This isn’t used as widely as Corsi.
PDO: Team shooting percentage + team save percentage. Sometimes referred to as “puck luck.”
This number generally hovers around 1000 (although a lot of places drop the last numeral so it’s just 100), and the stat itself hinges on the idea that everything eventually regresses to a mean (in this case, 100).
For instance, a team with a PDO of 104 (like the Rangers?) is likely not as good as they appear (which the Rangers are not), and a team at 98 is probably not as bad as they appear. Discrepancies like that are generally considered unsustainable. A really good goalie (like Lundqvist) can inflate this number a lot by inflating the team save percentage.
Why we look at this: PDO isn’t a good predictor, but it is a good indication of whether or not what your team is doing is sustainable. The Rangers, for instance, are unlikely to continue to experience the success they’ve had in the early season because their PDO is so high.
Scoring Chance: Not just a shot on goal, a scoring chance comes from a predefined area, also called the “high danger zone.” A higher percentage of shots taken from this area result in goals. A scoring chance doesn’t necessarily have to be a shot on goal (a shot that would result in a goal without a goalie present), just anything thrown in the direction of the net. This is why the number of scoring chances and the number of shots on goal will be different in a game.
Why we look at this: it’s a pretty good indicator of what team’s gonna win a hockey game.
Scoring chance against: scoring chances that happen against your team. Why we look at this: If your team has a high percentage of scoring chances against, you might want to look at what your team is doing (or not doing) in front of the net. Suppressing shots in this area has been a weakness on the part of the Stars defense.
Save Percentage: The total number of saves a goalie makes divided by the total number of shots against.
Shooting Percentage: The total number of goals divided by the total number of shots for. Both of these stats can be used to measure individuals or teams.
ZS%: Offensive Zone Start % (abbreviated as ZSO% on War-on-ice). This is an individual stat, measured by how many offensive faceoffs a player is present for divided by total number of offensive and defensive faceoffs they’re on the ice for.
This is primarily used to determine an individual player’s usage. If a player has a higher than 50% ZS%, it means the coach considers the player to be generally more offensively minded. Defensemen with higher ZS% can also be said to be “protected,” as the primary role for a defenseman in the offensive zone is to keep the puck in the zone. (*coughAaronEkbladcough*) A player with a lower ZS% is considered by their coach to be more defensively minded, a “shut down” player. Players with lower ZS% will probably also have a lower CF%.
Graphs explained!
Scoring Chances (per 60) through 12/6. League avg is starting to move closer to the historical avg. pic.twitter.com/ul8DdICFtj
— Carolyn Wilke (@Classlicity) December 7, 2015
This graph explained: the y axis (the vertical one) is scoring chances against inverted. That means that a team high up on that axis, like New Jersey, is good at shot suppression in their defensive zone. A team closer to the x axis (the horizontal one) faces a lot of high danger shots. The x axis is scoring chances for. A team further to the left on this graph doesn’t get a lot of scoring chances. A team further to the right gets more scoring chances.
Dull: a team that plays a game without a lot of scoring chances for either side
Fun: a team that plays a game with lots of scoring chances for both sides (fun is a subjective term here)
Bad: a team that faces more scoring chances against than they get scoring chances for
Good: a team that gets more scoring chances for than they face against
Scoring Chances v PDO through the weekend. pic.twitter.com/1nv1UB8OId
— Carolyn Wilke (@Classlicity) December 7, 2015
This graph explained: the y axis (the vertical one) is PDO (shooting % + save %). The x axis is scoring chances for.
Bad, lucky: A team with high PDO (puck luck) but not a lot of scoring chances
Bad, unlucky: A team with low PDO and not a lot of scoring chances
Good, unlucky: A team with low PDO but a lot of scoring chances
Good, lucky: A team with high PDO and a lot of a scoring chances
You would expect a good, unlucky team to regress to 100 and continue to win games. But a bad, lucky team (like the Rangers) will also regress to 100, and without the underlying “good” metrics (in this case, scoring chances for) will most likely start to lose more games. In other words, what the Dallas Stars are doing is sustainable, what the New York Rangers are doing is likely not.
This graph explained: This graph is basically what Carolyn did with the Scoring Chances chart above, just expressed differently, and individual players instead of teams. The y axis (vertical one) is shots against inverted, and the x axis is shots for. The same designations for what constitutes a good, bad, fun, or dull team applies to players here. The Stars were abysmal this game, despite winning it, which is why literally all of them appear on the “bad” side of the chart. The shots were 30 to 20 so the Stars didn’t precisely cover themselves in glory.
This graph explained: The thing I’ve always appreciated about Micah’s graphs is how visual they are. This is meant to show possession, it’s a visual representation of each player’s Corsi% when playing against particular players.
Blue for the Stars, red for the Canes, the amount of each color in the box says who had the puck more often while the players were on the ice together. The size of the box indicates how much time they spent on the ice together.
Bigger boxes mean a lot of time together, smaller boxes (or no boxes) mean less time together.
The opacity of the box indicates how many events there were. Brighter boxes for high event shifts, lighter boxes for low event shifts.
I hope this primer has been useful to you. Be sure to let me know if I’ve missed something you’ve been confused by!